DMD 28829 A Comparison of Pharmacokinetics Between …dmd.aspetjournals.org/content/dmd/early/2009/11/12/dmd.109.028829.full.pdfpredict human pharmacokinetics. Chou et al., reported

DMD 28829

1

A Comparison of Pharmacokinetics Between Humans and Monkeys

Takafumi Akabane, Kenji Tabata, Keitaro Kadono, Shuichi Sakuda, Shigeyuki Terashita,

Toshio Teramura

Analysis & Pharmacokinetics Research Labs

Discovery Drug Metabolism & Pharmacokinetics (TA, KT, KK, SS, ST, TT)

Astellas Pharma Inc.

21 Miyukigaoka, Tsukuba-city, Ibaraki , 305-8585 Japan

DMD Fast Forward. Published on November 12, 2009 as doi:10.1124/dmd.109.028829

Copyright 2009 by the American Society for Pharmacology and Experimental Therapeutics.

This article has not been copyedited and formatted. The final version may differ from this version.DMD Fast Forward. Published on November 12, 2009 as DOI: 10.1124/dmd.109.028829

at ASPE

T Journals on A

pril 1, 2020dm

d.aspetjournals.orgD

ownloaded from

http://dmd.aspetjournals.org/

DMD 28829

2

Running title page a) Species differences between humans and monkeys

b) Corresponding author

Takafumi Akabane

Analysis & Pharmacokinetics Research Labs

Discovery Drug Metabolism & Pharmacokinetics

Astellas Pharma Inc.

21 Miyukigaoka, Tsukuba-city, Ibaraki, 305-8585 Japan

Phone: +81-29-863-7010

Fax: +81-29-852-2972

E-mail: [email protected]

c) Number of

Text pages:52

Tables: 5

Figures: 3

References: 40

words in Abstract:250

words in Introduction:553

words in Discussion:1301

d) Abbreviations

AMI: amitriptyline

AUC: area under the plasma concentration-time curve

CLh: hepatic clearance

CLint: intrinsic clearance

CLt: total clearance

CYP: cytochrome P-450

DEX: dexamethasone

DIG: digoxin

F: bioavailability

Fa: fraction absorbed

fb: unbound drug fraction in blood

fe: urinary excretion ratio of unchanged

Fg: intestinal availability

Fh: hepatic availability

TAC: tacrolimus


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

3

fp: unbound drug fraction in plasma

HT: hydrochlorothiazide

i.v.: intravenous

IBU: ibuprofen

ke: disappearance rate constant

LC: liquid chromatography

Li: lithium

MDZ: midazolam

MS/MS: mass spectrometry

NIF: nifedipine

PAMPA:parallel artificial membrane permeability assay

Papp: apparent permeability

PBS: Phosphate-buffered saline

P-gp: p-glycoprotein

PRO: propranolol

Qh: blood flow rate in the liver

QID: quinidine

Rb: blood to plasma concentration ratio

TIM : timolol

VER : verapamil


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

4

Abstract

To verify the availability of pharmacokinetic parameters in cynomolgus monkeys, hepatic

availability (Fh) and the fraction absorbed multiplied by intestinal availability (FaFg) were

evaluated to determine their contributions to absolute bioavailability (F) after intravenous

and oral administrations. These results were compared with those for humans using 13

commercial drugs for which human pharmacokinetic parameters have been reported. In

addition, in vitro studies of these drugs, including membrane permeability, intrinsic

clearance, and p-glycoprotein affinity, were performed to classify the drugs on the basis of

their pharmacokinetic properties. In the present study, monkeys had a markedly lower F

than humans for 8 out of 13 drugs. While there were no obvious differences in Fh between

humans and monkeys, a remarkable species difference in FaFg was observed. Subsequently,

we compared the FaFg values for monkeys with the in vitro pharmacokinetic properties of

each drug. No obvious FaFg differences were observed between humans and monkeys for

drugs that undergo almost no in vivo metabolism. In contrast, low FaFg were observed in

monkeys for drugs that undergo relatively high metabolism in monkeys. These results

suggest that first-pass intestinal metabolism is greater in cynomolgus monkeys than in

humans, and that bioavailability in cynomolgus monkeys after oral administration is

unsuitable for predicting pharmacokinetics in humans. Additionally, a rough correlation was

also observed between in vitro metabolic stability and Fg in humans, possibly indicating the


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

5

potential for Fg prediction in humans using only in vitro parameters after slight

modification of the evaluation system for in vitro intestinal metabolism.


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

6

Introduction

Since the development of new drugs is a cost- and labor-intensive task, the selection of

candidates with good pharmacokinetic profiles is becoming increasingly common. This

practice minimizes the number of drug candidates dropped due to pharmacokinetic

problems during the clinical phase (Wishart, 2007)

When predicting human pharmacokinetics, the fraction absorbed (Fa), intestinal availability

(Fg), and hepatic availability (Fh) are the main factors to consider. Fh prediction has

become considerably accurate since several mathematical prediction models have been

established, including the physiological model, well-stirred model, parallel tube model, and

dispersion model (De Buck et al., 2007; Iwatsubo et al., 1996; Naritomi et al., 2001). For

FaFg, however, no quantitative prediction method has ever been established, although

several qualitative prediction methods using human intestinal microsomes have been

reported (Chiba et al., 1997; Fagerholm, 2007; Fisher and Labissiere, 2007; Shen et al.,

1997; Yang et al., 2007). For these reasons, we have used mainly animal pharmacokinetic

parameters to predict human FaFg in the drug discovery stage.

It has been regarded as natural that monkey metabolism is the most similar to that of

humans, and cynomolgus monkeys have been widely used in pharmacokinetic or

drug-safety studies for that reason. In the last decade, however, cynomolgus monkeys have


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

7

often been found to have a poorer bioavailability (F) than other animal species for many

compounds ( Tabata et al. 2009).

Recently several reports have stated that the intestinal transit process, namely Fa or Fg, is

major contributor to the low F in cynomolgus monkeys (Sakuda et al., 2006; Takahashi et

al., 2008). However, unlike Fh, which can be easily calculated via conventional PK analysis,

Fa and Fg are difficult to evaluate separately, particularly in the intestine. Consequently,

few systemic studies have explored the usefulness of using monkey FaFg parameters to

predict human pharmacokinetics.

Chou et al., reported that the Fa and total clearance, corrected by hepatic blood flow rate,

correlated well between humans and monkeys (Chiou and Buehler, 2002). This finding

suggests that the species difference may be caused by Fg. In addition, our laboratory

reported that midazolam (MDZ) had a markedly lower F (2.0%) in cynomolgus monkeys

than in humans (24-46%) , which was caused by high first-pass intestinal metabolism

(Sakuda et al., 2006). Similar results reported by Nishimura et. al. 2007 showed that

extensive metabolism in the intestine is the cause of MDZ’s low F in cynomolgus

monkeys.

In the present study, the following studies were performed to further investigate the species

differences between humans and cynomolgus monkeys. Thirteen commercially available

drugs for which the human pharmacokinetic parameters are known were selected and


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

8

classified into 5 categories according to cytochrome P-450 (CYP) isozyme selectivity and

p-glycoprotein (P-gp) affinity.

The 13 drugs were intravenously and orally administered to cynomolgus monkeys to obtain

the in vivo pharmacokinetic parameters (F, Fh, and FaFg) for each drug, which were then

compared with those in humans. In addition, we also obtained the in vitro parameters for all

13 drugs, including protein binding, blood/plasma concentration ratio (Rb), membrane

permeability, in vitro intrinsic clearance (CLint) in liver microsomes (CLint liver), CLint in

intestine microsomes (CLint intestine), and P-gp affinity.

In this report, we discuss the main factor affecting the species difference between humans

and cynomolgus monkeys indicated by these results, and the adequacy of cynomolgus

monkeys as an animal model for predicting human pharmacokinetics.


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

9

Materials and Methods

Chemicals

MDZ (Dormicam, 5 mg/mL solution for intravenous injection) was obtained from Astellas

Pharma Inc. (Tokyo, Japan). Tacrolimus (TAC) synthesized at our laboratory was used.

Lithium carbonate (Li) was purchased from Kanto Chemical Co., Inc. (Tokyo, Japan).

Hydrochlorothiazide (HT), verapamil (VER), propranolol (PRO), and amitriptyline (AMI)

were purchased from Wako Pure Chemical (Osaka, Japan). Dexamethasone (DEX),

nifedipine (NIF), quinidine (QID), timolol (TIM), and ibuprofen (IBU) were purchased

from Sigma Chemicals (St. Louis, MO, USA). Liver and intestine microsomes from humans

and cynomolgus monkeys were purchased from Xenotech (Lenexa, KS, USA). All other

reagents and solvents were commercial products of analytical grade.

Animals

Male cynomolgus monkeys (Shin Nippon Biomedical Laboratories, Ltd., Kagoshima,

Japan, Astellas Research Technology, Osaka, Japan) weighing about 5 kg were used. The

animal experiment was conducted according to the ethical rules of each company.

Selected drugs and categorization

We allocated the 13 drugs into 5 categories (Type A – E), according to their


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

10

pharmacokinetic properties in humans, as follows: membrane permeability, CYP isozyme

selectivity, and P-gp affinity (Kivist et al., 2004; Yang et al., 2006; Yu, 1999) (Table 1).

Type A:

The drugs categorized as type A are indicator drugs that undergo no metabolism in humans

and are not P-gp substrates. For each of these, almost all of the absorbed drug is excreted

into urine as the unchanged form. Li, which has a high F in humans (94.5%: Arancibia et al.,

1986), and HT, which has a moderate F in humans (60.2%: Patel et al., 1984), were

assigned to this category.

Type B:

The drugs categorized as type B are CYP3A4 substrates, and have very weak, if any,

affinity for P-gp.

DEX, which has a high F in humans (81.4%: Duggan et al., 1975), NIF, and MDZ, which

have a moderate F in humans (41.2%: (Holtbecker et al., 1996) and 30.0%: (Thummel et al.,

1996) respectively) were assigned to this category.

Type C:

The drugs categorized as type C are substrates of both CYP3A4 and P-gp.


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

11

QID, which has a high F in humans (79.5%: Greenblatt et al., 1977):, as well as TAC and

VER, which have a moderate F in humans (23.3%: (Moller et al., 1999) and 18.0%:

(McAllister and Kirsten, 1982), respectively) were assigned to this category.

Type D:

DIG, which is substrate of P-gp, but not CYP3A4, was categorized as type D. DIG has a

high F in humans (65.3%: Hinderling and Hartmann, 1991) and undergoes almost no

metabolism in the human body, ie, it undergoes only P-gp efflux during the absorption

process in the intestine.

Type E

The drugs categorized as type E are mainly metabolized by the CYP isozyme (except 3A4)

and have very weak, if any, affinity for P-gp.

IBU and TIM, which have a high F in humans (100%: (Martin et al., 1990), 61.0%:

(Wilson et al., 1982), respectively):, as well as AMI and PRO, which have a moderate F in

humans (47.7%: (Schulz et al., 1983), and 29.0%: (Borgstrom et al., 1981), respectively)

were assigned to this category.

See “Table 1” for CYP isozymes which contribute to each drug metabolism.


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

12

Pharmacokinetic study in cynomolgus monkeys

Intravenous and oral administrations were performed with a washout period of at least 7

days between each type of administration. Animals were fasted for approximately 17 h

before dosing. Blood samples were collected from the antebrachial vein, kept in an

ice-water bath, and then centrifuged at 10,000 rpm for 1 min at 4 °C. The plasma samples

were kept in a deep freezer (approximately -20 °C) until analysis. The experimental

conditions for the pharmacokinetic studies, including doses, dosing solution, dosing volume,

and sampling time for each drug, are shown in Table 2.

Values obtained from the literature were used as the pharmacokinetic parameter values for

all selected drugs in humans as well as those for MDZ in cynomolgus monkeys.

Measurement of model compounds plasma concentration in cynomolgus monkeys

The concentrations of model drugs in cynomolgus monkey plasma were determined using

atomic absorption, enzyme immunoassay analysis, or high-performance liquid

chromatography (LC) coupled with tandem mass spectrometry (MS/MS) with sample

pre-treatment).

Atomic absorption method

Lithium

The lithium level in the plasma was determined using atomic absorption in accordance with


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

13

the method of Pybus and Bowers, 1997.

Enzyme immunoassay analysis

Dexamethasone and Tacrolimus

The DEX level in the plasma and the TAC level in the blood were determined using enzyme

immunoassay.

After extracting as follows, an aliquot was used as the sample for analysis by enzyme

immunoassay (Tamura et al., 1987).

A 50-uL aliquot of plasma was buffered with 1% skim milk/PBS. After the addition of 1 mL

of distilled water, the mixture was extracted with 5 mL of diethyl ether, and the solvent was

removed under a stream of nitrogen gas. The residue was then dissolved in 250 uL of 1%

skim milk/PBS.

LC-MS/MS analysis

The plasma concentrations of all other drugs were determined using LC-MS/MS.

The LC-system comprised a LC-VP/LC-10A series (Shimadzu, Kyoto, Japan) or HP-1100

series HPLC (Agilent Technology Inc., Santa Clara, CA, USA). The MS/MS experiments

were conducted using API-2000 or API-3000 LC/MS/MS systems (Applied Biosystems,

Foster, CA, USA).


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

14

The details of the LC-MS/MS conditions, including the machines and columns used for

each drug, are shown in Table 3.

Hydrochlorothiazide

A 200-uL aliquot of plasma was buffered with 500 uL of 10 mM phosphate buffer adjusted

to pH 3.0. After adding 100 uL of acetonitrile and 20 uL of internal standard solution (1

ug/mL of diclofenac in 50% acetonitrile), the mixture was extracted with 4 mL of ethyl

acetate, and the solvent was removed under a stream of nitrogen gas.

Then, the residue was dissolved in 100 uL mobile phase, and a 40-uL aliquot was injected

into the LC-MS/MS (molecular>product: m/z = 296>269 [M+H]-).

Nifedipine

A 50-uL aliquot of plasma, 50 uL of 50% acetonitrile, and 100 uL of internal standard

solution (1 ug/mL of in house compound A in acetonitrile) were mixed well, and then

centrifuged to remove precipitated protein. The supernatant (100 uL) was then decanted,

and 30 uL was injected into the LC-MS/MS (molecular>product: m/z = 347>315 [M+H]+).

Quinidine, Verapamil, Propranolol, Amitorptyrine, and Timolol

A 200-uL aliquot of plasma was buffered with 500 uL of saturated sodium bicarbonate


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

15

solution. After the addition of 50 uL acetonitrile and 50 uL of internal standard solution (1

ug/mL of in house compound B in 50% acetonitrile), the mixture was extracted with 3 mL

of tertiary butyl methyl ether, after which the solvent was removed under a stream of

nitrogen gas.

The residue was then dissolved in 200 uL mobile phase, and a 20-uL aliquot was injected

into the LC-MS/MS (molecular>product: QID m/z = 325>307 [M+H]+, VER m/z =

455>165 [M+H]+, TIM m/z = 317>261 [M+H]+, AMI m/z = 278>117 [M+H]+, PRO m/z =

260>116 [M+H]+)

Digoxin

A 200-uL aliquot of plasma was buffered with 500 uL of 10 mM phosphate buffer adjusted

to pH 3.0. After the addition of 100 uL of acetonitrile and 50 uL of internal standard

solution (1 ug/mL of digitoxin in 50% acetonitrile), the mixture was extracted with 3 mL of

ethyl acetate, and the solvent was removed under a stream of nitrogen gas.

The residue was then dissolved in 100 uL mobile phase, after which a 20-uL aliquot was

injected into the LC-MS/MS (molecular>product: m/z = 798>391 [M+NH4]+) .

Ibuprofen

A 200-uL aliquot of plasma was buffered with 500 uL of 5 mM phosphoric acid. After the


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

16

addition of 50 uL of acetonitrile and 50 uL of internal standard solution (1 ug/mL of

diclofenac in 50% acetonitrile), the mixture was extracted with 3 mL of tertiary butyl

methyl ether ether, and the solvent was removed under a stream of nitrogen gas.

The residue was then dissolved in 200 uL of mobile phase, and a 20-uL aliquot was

injected into the LC-MS/MS (molecular>product: m/z = 205>161, [M+H]-).

Blood-to-plasma concentration ratio (Rb)

One milliliter of human and cynomolgus monkey blood was spiked with 10 μL of 100

μg/mL standard solution (1,000 ng/mL final) and preincubated in a shaking water bath at

37 °C for 10 min. A 200-μL aliquot was then analyzed to determine the drug concentration

in the blood. The remaining samples were centrifuged at 4°C and 1,800 × g for 10 min, after

which the drug concentration in 200-μL aliquots of plasma were determined. The Rb was

then calculated from the concentrations of drug per milliliter of blood and plasma.

All data regarding TAC level in humans and cynomolgus monkeys were determined by

blood level base because the Rb value of TAC has been reported to be non-linear, with

values between 10 and 40 depending on the drug concentration in humans (Wallemacq et al.,

1993).

Parallel artificial membrane permeability assay (PAMPA)


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

17

The PAMPA method was carried out using a PAMPA Evolution instrument from pION

INC. (Woburn, MA, USA) (Avdeef et al., 2005).

The lipid solution consisted of a 20% (w/v) dodecane solution and lecithin mixture. The

donor solutions consisted of test compounds dissolved in dimethylsulfoxide (10 mM)

diluted in pH 6.5 buffer (final concentration of 50 uM). The acceptor plate was filled with

1% (w/v) SDS in water, and the pH was adjusted to 7.4 with 1N hydrochloric acid. The test

plate was incubated for 120 min at 30 °C. The concentration of each test compound in the

reference, donor, and acceptor plates was measured with a UV plate reader. The

permeability coefficient was calculated using Evolution Library Manager software V2.2

(pION INC., Woburn, MA, USA).

Plasma protein binding

The plasma protein binding (unbound drug fraction in plasma) was determined using the

Equilibrium dialysis method or ultracentrifugation method and the following equations:

Protein binding (%) = (1-fp) × 100 Equation 1

fp = concentration in filtrate or supernatant/concentration in serum Equation 1’

Where fp is the unbound drug fraction in plasma. The unbound drug fraction in blood (fb)

was calculated by dividing fp by Rb.


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

18

Equilibrium dialysis method

A Diachema SC-101-M1OH (DIANORM) dialyzing membrane, which is impermeable to

substances with molecular weights greater than 10,000, was used.

Aliquots (3.5-mL) of human and cynomolgus monkey plasma were spiked with 35 μL of

100 μg/mL standard solution (1,000 ng/mL final) and pre-incubated in a 37 °C shaking

water bath for 10 min.

One milliliter of mixture and isotonic phosphate buffer solution (pH 7.4 ) was put into the

dialyzing cell and receptor cell, respectively. After 4 h of incubation at 37 °C, the plasma

mixture and buffer sample were stored in 100-uL aliquots at -20 °C until analysis.

Ultracentrifugation method

Ten microliters of standard solution (100 μg/mL) were added to 1,000 μL of human or

cynomolgus monkey plasma. The calibration samples were prepared by adding 17 μL of

50% acetonitrile to 1,700 μL of human or cynomolgus monkey plasma. These samples were

then centrifuged and 436,000 × g for 140 min at 37 °C using a Beckman Optimal TL

ultracentrifuge (Beckman Coulter, Fullerton, CA, USA). After ultracentrifugation, the

unbound fp was calculated by dividing the concentration of drugs in the supernatant by that

in the plasma.


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

19

In Vitro Metabolism in liver and intestine microsomes

Metabolism study conditions

The time courses of the unchanged drugs were obtained. Each drug was incubated at 37 °C

with a reaction mixture (1 mL) containing 500 μL of 200 mM potassium-phosphate buffer

(pH 7.4), 100 μL of 1 mM EDTA-NaOH (pH 7.4), 100 μL of liver or intestine microsomes

solution (the final concentration of microsomal protein was 0.05 mg/mL for TAC, 0.5

mg/mL for HT and DIG, and 0.2 mg/mL for all other drugs), 190 μL of distilled water, and

10 μL of each compound solution in 50% acetonitrile (final concentration: 0.2 μM).

After 5 min of preincubation, the reaction was initiated by the addition of 100 uL of a

NADPH-generating system.

The reaction was terminated by adding 100 uL of reaction mixture to 200 μL of acetonitrile

including the internal standard at various time periods.

After stopping the enzyme reaction, the reaction mixture of TAC and DIG were extracted

with 3 mL of tertiary butyl methyl ether, and the solvent was removed under a stream of

nitrogen gas.

The residue was then dissolved in 150 uL of mobile phase, and a 10-uL aliquot was

injected into the LC-MS/MS.

The reaction mixture of DEX and NIF were centrifuged at 10,000 × g for 5 min. The

supernatant (100 uL) was then decanted, and 30-uL aliquots was injected into the


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

20

LC-MS/MS.

The reaction mixtures of all other drugs were centrifuged at 10,000 × g for 5 min, The

supernatants (100 uL) were decanted, and a 10-uL aliquots were injected into the

LC-MS/MS.

In this experiment, the unchanged concentrations of all drugs were determined using

LC-MS/MS analysis. Mass number of molecular ion and product ion for each compounds

were identified as fellows (polarity, molecular>product ): HT m/z = 296>269 [M+H]-, DEX

m/z = 393>91 [M+H]+, NIF m/z = 347>315 [M+H]+, MDZ m/z = 326>291 [M+H]+, QID

m/z = 325>307 [M+H]+, TAC m/z = 821>769 [M+NH4]+, VER m/z = 455>165 [M+H]+,

DIG m/z = 780>85 [M+H]-, IBU m/z = 205>161 [M+H]-, TIM m/z = 317>261 [M+H]+,

AMI m/z = 278>117 [M+H]+, PRO m/z = 260>116 [M+H]+.

The Prominence 2000 series (Shimadzu, Kyoto, Japan) was used as the LC-system. The

MS/MS analyses were conducted on an API-3200 LC-MS/MS system (Applied Biosystems,

Foster, CA). For TAC, an Alliance HT Waters 2790 separations module and Micromass

Quattro Ultima (Waters Corporation, Milford, MA, USA) were used for the LC-MS/MS

analysis.

The Supelco RP-Amide (3 μm, 3.0 x 31 mm; Supelco, Inc., Bellefonte, PA, USA) was

used as the analysis column for HT and DIG. The Capcell PAK MG (3 μm, 2.0 x 35 mm;

Shiseido Corporation, Kyoto, Japan) HPLC column was used for all other drugs.


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

21

The flow rate was 0.3 mL/min. The column temperature was 50 °C. The gradient system

was used, starting with an ammonium acetate concentration of 20 mM (pH 4.8)/acetonitrile

(9:1) for 0.5 min, and increasing the ratio of acetonitrile to 20 mM ammonium acetate (pH

4.8)/acetonitrile (1:9) over 0.5 min, which was then held for 2.5 min. The initial conditions

were restored over 0.1 min, after which the column was re-equilibrated for 1 min.

Calculation of in vitro intrinsic clearance in liver (CLint liver)

CLint liver was calculated using Equation 2 based on the time course of the residual ratio of

the unchanged drugs as determined using least squares linear regression (Naritomi et al.,

2001).

Clint liver (mL /min/mg protein) = ke / microsomal protein concentration Equation 2

where ke is the disappearance rate constant.

In the case of liver microsomes study, the units of CLint liver values were converted to per

kilogram of body weight using Equation 3.

CLint liver (mL/min/kg)

= CLint liver (mL/min/mg protein) x SF1 (mg protein/g liver) x SF2(g liver/kg body weight)

Equation 3


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

22

where SF1 is the microsomal protein content per gram of liver, (48.8 was used for both

species (Naritomi et al., 2001), assuming that the SF1 in cynomolgus monkeys is the same

as in humans) and SF2 is the liver weight per kilogram of body weight (25.7 and 30.0 were

used for humans and cynomolgus monkeys, respectively: Davies and Morris, 1993).

Calculation of in vitro intrinsic clearance in intestine (CLint intestine)

CLint intestine was calculated using Equation 2’ based on the time course of the residual ratio

of the unchanged drugs as determined using least squares linear regression (Naritomi et al.,

2001).

CLint intestine (uL/min/mg protein) = ke / microsomal protein concentration Equation 2’

P-gp ATPase-Assay

Each drug was dissolved in dimethylsulfoxide (0.1-100 μM final) and preincubated for

5min with 2 ug/mL of human P-gp membrane (BD Gentest, Woburn, MA, USA) in 50 mM

MES buffer (pH 6.8 adjusted with Tris) containing 2 mM EGTA, 2 mM dithiothreitol, 50

mM potassium chloride, and 5 mM sodium azide. Then, the ATPase reaction was started by

the addition of 50 mM Mg-ATP solution. After 20 min incubation at 37 °C, the reaction was

stopped by adding 20 uL of 10% sodium dodecyl sulfate containing Antifoam A.

Subsequently, 200 uL of ammonium molybdate / zinc acetate was added for color


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

23

development, and the mixture was incubated for another 20 min at 37 °C. After incubation,

the amount of liberated phosphate was measured using the UV absorption method (630 nm).

Baseline activity was determined by reading incubated 100 uM sodium orthovanadate.

Finally, ATPase activity was determined as the amount of liberated phosphate per milligram

protein per minute.

VER was evaluated in all ATPase assays, and the ATPase activity of each drug was

normalized by dividing by the VER ATPase activity for each experiment.

Calculation of in vivo pharmacokinetic parameters.

Plasma concentration data were analyzed individually at each point in time, and

pharmacokinetic parameters were calculated using a model-independent method. F, FaFg

and Fh were then calculated from these pharmacokinetic parameters and Rb (See

“Blood-to-plasma concentration ratio” in the Materials and Methods section) using the

formulas shown below. For Li and HT,, we assumed these drugs underwent almost no in

vivo metabolism and that their FaFg values (meaning Fa in this case) were equal to F.

The F values for the drugs in cynomolgus monkeys were determined using Equation 4.

F(%) = {AUCinf (p.o.) / AUCinf (i.v.)}･(Dose i.v. / Dose p.o.) × 100 Equation 4


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

24

where AUCinf (i.v.) and AUCinf (p.o.) are the area under the plasma concentration-time

curve calculated using the trapezoidal rule with extrapolation from the last measured plasma

concentration to infinity after intravenous and oral administrations, respectively.

The Fh of drugs were determined using Equation 5 and assuming that the elimination of

drugs from the body after intravenous administration consisted of liver metabolism and

renal excretion.

Fh = 1-{(CLh/Rb) / Qh}、CLh = CLt×(1-fe) Equation 5

fe = urinary excretion of unchanged of unchanged drug after intravenous administration.

Where Qh is the blood flow rate in the liver (the human and cynomolgus monkey Qh values

were 20.7 and 43.6 mL/min/kg, respectively: Davies and Morris, 1993), CLh is hepatic

clearance, CLt is total clearance, and fe is the urinary excretion ratio of the unchanged drug

after intravenous administration. In cases where the fe value was not available, the CLh was

assumed to be equal to the CLt.

The drug FaFg values were determined using Equation 6 assuming that the F was


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

25

expressed as the product of FaFg and Fh

F(%) = Fa×Fg×Fh×100, FaFg = {F(%)/100} /Fh Equation 6

The F, FaFg, and Fh values of each drug in humans were also calculated in a similar

manner using the reported pharmacokinetic parameters.


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

26

Result

Comparison of pharmacokinetic parameters between humans and cynomolgus

monkeys. The in vivo pharmacokinetic parameters, F, FaFg, and Fh for all 13 drugs are summarized

in Table 4.

Each drug’s cynomolgus monkey F, FaFg, and Fh values are plotted against those in

humans in Fig. 1.

Correlation of the F between humans and cynomolgus monkeys

The F values of all drugs observed in cynomolgus monkeys were compared with those in

humans. The results showed that the F value for Li, DEX and IBU in humans and

cynomolgus monkeys were similar, followed by HT and DIG were almost similar (< 2

fold).

In contrast, with the exception of DEX and IBU, many of the CYP substrate drugs had a

markedly lower F in cynomolgus monkeys than in humans.

Type A:

The F values for Li in humans and cynomolgus monkeys were similar (94.5% / 97.9%), and

HT showed slightly lower F values in cynomolgus monkeys (30.7%) than humans (60.2%).

Type B:

For DEX, the F values in humans and cynomolgus monkeys were similar (81.4% and


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

27

78.9%, respectively). However, the F values for NIF and MDZ in cynomolgus monkeys

were markedly lower (9.3% and 2.0% (Sakuda et al., 2006), respectively) than those in

humans (41.2% and 30.0%, respectively).

Type C:

The type C drugs, QID, TAC and VER, which are known to be substrates for both

CYP3A4 and P-gp in humans, had markedly lower F values (4.5, 0.5 and 0%, respectively)

in cynomolgus monkeys than in humans (79.5, 23.3, and 18.0%, respectively).

Type D:

The DIG, which is a typical substrate of P-gp had a slightly lower F value in cynomolgus

monkeys (45.0%) than humans (65.3%). This finding was similar to that for HT.

Type E:

While IBU’s F value was almost the same in both species, those for TIM, AMI, and PRO

were lower in cynomolgus monkeys (10.8, 1.3 and 3.3%) than in humans (61.0, 47.7, and

29.0%). These findings were similar to those for Type B drugs.

No significant correlation between the CYP isozyme selectivity of drugs and their F values

in cynomolgus monkeys was observed.


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

28

Correlation of the Fh between humans and cynomolgus monkeys

The correlations between the human and cynomolgus monkey Fh values for the 13 drugs

are shown in Fig. 1: Fh.

The Fh values in cynomolgus monkeys were similar to those in humans for all drugs except

VER (Fh was calculated as 0 in cynomolgus monkeys), because the plots for the drugs were

the same or nearly the same (Fig. 1: Fh and Table 4).

Li and HT underwent almost no in vivo metabolism; therefore, the Fh values were

considered to be 1.

Correlation of the FaFg between humans and cynomolgus monkeys

As shown in Fig. 1: FaFg, the FaFg values for Li, DEX, and IBU were similar in both

humans and cynomolgus monkeys (0.95/0.98, 0.93/0.85, and 1/1, respectively).

For HT and DIG, the FaFg values in cynomolgus monkeys were slightly lower than those

in humans (0.60/0.31 and 0.67/0.48 in humans and cynomolgus monkeys, respectively).

For the other 7 drugs (except VER), the F in cynomolgus monkeys was low, and a markedly

low FaFg was observed. These tendencies correlated well with those of the F values

(assuming Fh = 1 for Li and HT, which means F = FaFg).

In vitro parameters


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

29

In this study, some additional in vitro assays were performed to evaluate the drugs’ (except

Li) in vitro pharmacokinetic properties. These assays included determination of the

blood-to-plasma concentration ratio, membrane permeability, in vitro metabolic stability

assay using human and cynomolgus monkey liver microsomes, plasma protein binding, and

P-gp affinity. The results are summarized in Table 5.

Membrane permeability

As shown in Table 5, almost all drugs except HT and DIG showed good membrane

permeability (>apparent permeability coefficient of more than 10). Taking the F values into

consideration, the HT and DIG were speculated to be absorbed moderately in cynomolgus

monkeys.

These results suggest that all tested drugs were well-absorbed or relatively well-absorbed

in cynomolgus monkeys, even though many drugs had a low F.

Metabolic stability in liver microsomes

For HT and DIG, no depletion was observed, and the intrinsic clearance for DEX MDZ,

and IBU in both humans and cynomolgus monkeys were almost the same (66/24

mL/min/kg, 877/1,422 mL/min/kg and 38/25 mL/min/kg, respectively).

Intrinsic clearance values for the other seven drugs were higher in cynomolgus monkeys


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

30

than in humans (Table 5). Although Fh correlated well between humans and cynomolgus

monkeys for all tested drugs except VER, these drugs were metabolized more rapidly in

cynomolgus monkey microsomes than in human microsomes. Further, the fb×CLint liver/Qh

for NIF, VER, PRO, and AMI were found to be higher (>4) after taking fb and blood flow

rate in the liver into consideration, indicating that these drugs might undergo rapid

metabolism in the livers of cynomolgus monkeys.

Metabolic stability in intestine microsomes

The CLint intestine was expressed by uL/min/mg protein because there is no widely used

physiological conversion model from uL/min/ mg protein to uL/min/kg in intestine. The

CLint intestine values for NIF, MDZ, QID, TAC, and VER in cynomolgus monkey were 612,

1635, 212, 4663, and 696 uL/min/mg protein, respectively. As well as in human, the values

were 138, 385, no depletion, 625, and 69 uL/min/mg protein for each (Fig.2 and Table 5). In

contrast, no significant decreases in other drugs were observed both in human and

cynomolgus monkey intestine microsomes.

ATPase assay

The ATPase activity of all drugs was normalized by dividing them by the VER value.

As shown in Table 5, the ATPase activity of QID, DIG, and TAC were higher than that of

VER. For PRO, AMI, TIM, and IBU, the ATPase activity values were similar to the VER


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

31

value, while the HT, DEX, NIF, and MDZ were lower. There was no significant correlation

between P-gp affinity and F values in cynomolgus monkeys observed.


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

32

Discussion

While cynomolgus monkeys are often used for pharmacokinetic studies for drug discovery,

it remains unclear whether this is a useful animal species for predicting human

pharmacokinetics.

In this study, we investigated the pharmacokinetic profile of 13 commercially available

drugs in cynomolgus monkeys and compared their pharmacokinetic parameters with those

in humans. The results showed that the majority of the drugs tested (8 out of 13) had a

markedly lower F in cynomolgus monkeys (<15%). We explored the reasons for these

species differences and suggest some possibilities as below.

Species differences in hepatic metabolism

The Fh values in humans and cynomolgus monkeys were almost the same for the 12 drugs

(except VER).

No obvious species difference were revealed for hepatic metabolism, regardless of CYP

isozyme selectivity.

These results suggested that the values obtained from cynomolgus monkeys after

intravenous administration were useful for predicting human pharmacokinetic parameters,

such as CLt or Fh.

These findings agreed with the consistency seen between the species with regard to CYP


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

33

isozyme amino acid sequence (over 90% agreement) (Uno et al., 2007)..

A species difference in Fh was apparent for VER, which was explained by the difference in

the rate of hepatic metabolism. The fb x CLint liver/Qh of VER in cynomolgus monkeys was

much higher than that in humans, which agreed with the in vivo observation.

Species differences in the intestinal transit process

The fact that all drugs with a low F in cynomolgus monkeys had low FaFg values

indicates that the low FaFg is attributable to the low F, in cynomolgus monkeys

specifically.

The FaFg values for Li, DEX, and IBU were correlated well between humans and

cynomolgus monkeys. The common properties of these 3 drugs are as follows: 1) they have

good membrane permeability (Li is absorbed via a paracellular pathway), 2) they are not

P-gp substrates, and 3) they undergo little or no in vivo metabolism (see Tables 4 and 5).

Subsequently, the FaFg correlation between humans and cynomolgus monkeys was found

to be weak for both HT and DIG. The FaFg values for these drugs in cynomolgus monkeys

was slightly lower than that in humans. The common properties of these 2 drugs are as

follows: 1) they have moderate membrane permeability, and 2) they undergo almost no in

vivo metabolism (Tables 4 and 5). Although HT is not a P-gp substrate, DIG was found to

cause high activity in the ATPase assay.


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

34

These results suggest that membrane permeability and P-gp efflux are partial contributors

to the low F in cynomolgus monkeys.

In contrast, the other 7 drugs (except VER), which had a markedly low FaFg in

cynomolgus monkeys, were metabolized by CYP enzymes and had relatively high CLint liver

or CLint intestine values in cynomolgus monkeys. These drugs also showed good membrane

permeability (Table 5).

These findings suggest the possibility that these drugs were extensively metabolized in the

cynomolgus monkey intestine, and the low FaFg was caused by intestinal metabolism

rather than poor absorption. In fact, all of five drugs, which observed good FaFg correlation

in both species, undergo little or no in vivo CYP metabolism.

The major species difference factor between humans and cynomolgus monkeys

There have been several reports that focused on the species differences between humans

and monkeys (Chiou et al., 2002; Sakuda et al., 2006; Takahashi et al., 2008). The present

study, however, showed that drugs that satisfy properties listed below have similar FaFg or

F values in both humans and cynomolgus monkeys.

1) Good membrane permeability

2) Not a P-gp substrate

3) Undergoes little or no in vivo metabolism

In contrast, drugs that are CYP substrates and are relatively or rapidly metabolized in


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

35

cynomolgus monkeys could have markedly low F values because of their low FaFg values,

even if the drugs have a low CLt.

The potential reasons for these findings are as follows: 1) the amount of CYP enzyme

expressed in cynomolgus monkey intestine is higher than that in humans, even though

CYP3A4 is major intestinal enzyme in humans, and 2) the enzyme expressed in

cynomolgus monkey intestine has higher activity (Vmax/Km) than that in humans. In

order to make clearly understand these speculations, additional in vitro studies using

intestine microsome were conducted with same condition as liver microsomes study. In

cynomolgus monkey, the values of CLint intestine for NIF, MDZ, QID, TAC, and VER were

612, 1635, 212, 4663, and 696 uL/min/mg protein, respectively. As well as in human, the

values were 138, 385, no depletion, 625, and 69 uL/min/mg protein for each. These five

compounds which have low F in cynomolgus monkey showed markedly larger values in

cynomolgus monkey than those in human (Fig.2). In contrast, no significant decreases in

other drugs were observed both in human and cynomolgus monkey intestine microsomes.

While the cynomolgus monkey CYP isozyme corresponding to human CYP3A4 is

CYP3A8 (Uno et al., 2007)., it is unclear whether CYP3A8 is also major enzyme in the

cynomolgus monkey intestine. In fact, a lower FaFg in cynomolgus monkeys was also

observed for Type E drugs (mainly metabolized by CYP 2C9, 2C19, or 2D6).

Although it is possible that glucuronide conjugates contributed to the low F obtained for


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

36

PRO (Walle et al., 1979)., further studies are need to explain this observation.

Since all drugs with a low F in cynomolgus monkeys show good membrane permeability

in the present study, first-pass intestinal metabolism must be the most critical factor

affecting species differences between humans and cynomolgus monkeys.

We also investigated the pharmacokinetics of several drugs in rats and/or dogs, and the

FaFg in rats or dogs correlates better with humans than cynomolgus monkeys (Tabata et

al., 2009). Further studies are needed to clarify the species differences for FaFg, including

the contribution of permeability, intestinal first-pass metabolism, and P-gp excretion.

The usability of cynomolgus monkey pharmacokinetic parameters for predicting

pharmacokinetic in humans

These results suggest that a go/no go decision does not have to be made immediately, even

if a candidate has a markedly low F in cynomolgus monkeys. In such cases, the main factor

causing low F in cynomolgus monkeys may be evaluated separately from Fa, Fg, and Fh. If

the cause is found to be Fg, the candidate could still have an acceptable pharmacokinetic

profile in humans.

Recognition of the importance of intestinal metabolism has increased over recent years.

Many studies using intestinal microsomes are in progress in our laboratory in an attempt to

establish a system for evaluating human Fg.


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

37

Interestingly, a rough correlation was observed between CLint liver and Fg in humans (Fig.

3) in this study, indicating the possibility that Fg prediction in humans using only in vitro

parameters may be possible with slight but elaborated modification of the evaluation system

for in vitro intestinal metabolism. In fact, when evaluation of intestinal metabolism was

inadequate, we successfully predicted the human pharmacokinetics for several in-house

candidate drugs with a markedly low F in cynomolgus monkeys by using human in vitro

parameters for each candidate, including membrane permeability, metabolic stability in liver

microsomes, and P-gp affinity (in-house data). These low values for F in cynomolgus

monkeys were virtually thought to be due to low Fg.

In conclusion, many drugs had a markedly low F in cynomolgus monkeys despite having

relatively good F in humans. These findings are speculated to be attributable mainly to

first-pass intestinal metabolism. Consequently, the pharmacokinetic parameters obtained for

a candidate after oral administration to cynomolgus monkeys are not adequate for directly

predicting human pharmacokinetics.

The accurate prediction of Fg in humans eventually becomes necessary to predict human

pharmacokinetics with more accuracy. And the slight but elaborated modification of the

evaluation system for in vitro intestinal metabolism, which is under development in our

laboratory (Kadono et al., 2007), may enable us to estimate the Fg in humans, subsequently

it becomes possible to predict accurate human pharmacokinetics in the near future.


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

38

【References】

Arancibia A, Corvalan F, Mella F and Concha L. (1986) Absorption and disposition

kinetics of lithium carbonate following administration of conventional and

controlled release formulations. Int J Clin Pharmacol Ther Toxicol 24:240-245.

Avdeef A, Artursson P, Neuhoff S, Lazorova L, Grasjo J and Tavelin S. (2005)

Caco-2 permeability of weakly basic drugs predicted with the double-sink PAMPA

pKa(flux) method. Eur J Pharm Sci. 24:333-349

Borgstrom L, Johansson CG, Larsson H and Lenander R.（1981) Pharmacokinetics

of propranolol. J Pharmacokinet Biopharm. 9:419-429.

Chiba M, Hensleigh M and Lin JH. (1997) Hepatic and intestinal metabolism of

indinavir, an HIV protease inhibitor, in rat and human microsomes. Major role of

CYP3A. Biochem. Pharm. 53:1187-1195

Chiou WL and Buehler PW. (2002) Comparison of oral absorption and

bioavailablity of drugs between monkey and human. Pharm Res. 19:868-74

Davies B and Morris T (1993) Physiological parameters in laboratory animals and

humans. Pharma. Res. 10:1093-1095

De Buck SS, Sinha VK, Fenu LA, Nijsen MJ, Mackie CE and Gilissen RA. (2007)

Prediction of human pharmacokinetics using physiologically based modeling: a


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

39

retrospective analysis of 26 clinically tested drugs. Drug Metab.Dispos.

35:1766-1787

Duggan DE, Yeh KC, Matalia N, Ditzler CA and McMahon FG. (1975)

Bioavailability of oral dexamethasone. Clin Pharmacol Ther. 18:205-209.

Evans GH, Nies AS and Shand DG. (1973) The disposition of propranolol. 3.

Decreased half-life and volume of distribution as a result of plasma binding in man,

monkey, dog and rat. J Pharmacol Exp Ther. 186:114-122

Fagerholm U. (2007) Prediction of human pharmacokinetics--gut-wall metabolism.

J Pharm Pharmacol. 59:1335-1343

Fisher MB and Labissiere G. (2007) The role of the intestine in drug metabolism

and pharmacokinetics: an industry perspective. Curr Drug Metab. 8:694-699

Greenblatt DJ, Pfeifer HJ, Ochs HR, Franke K, MacLaughlin DS, Smith TW and

Koch-Weser J. (1977) Pharmacokinetics of quinidine in humans after intravenous,

intramuscular and oral administration. J Pharmacol Exp Ther. 202:365-378

Hinderling PH and Hartmann D. (1991) Pharmacokinetics of digoxin and main

metabolites/derivatives in healthy humans. Ther Drug Monit. 13:381-401


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

40

Holtbecker N, Fromm MF, Kroemer HK, Ohnhaus EE and Heidemann H.

(1996)The nifedipine-rifampin interaction. Evidence for induction of gut wall

metabolism. Drug Metab Dispos. 24:1121-1123

Iwatsubo T, Hirota N, Ooie T, Suzuki H and Sugiyama Y. (1996) Prediction of in

vivo drug disposition from in vitro data based on physiological pharmacokinetics.

Biopharm. Drug Dispos. 17:273-310

Kadono K, Akabane T, Tabata K, Mitani Y, Hirabayashi H, Miura H and Teramura

T. (2007) An Empirical ADME Screening System for Drug Discovery in Astellas

(2) -A Simplified Prediction Method of Human Intestinal Availability Suitable for

the Early Stage of Drug Discovery-. Drug Metabolism Reviews, 39 S1:121-122

Kivist KT Niemi M and Fromm MF. (2004) Functional interaction of intestinal

CYP3A4 and P-glycoprotein. Fundam Clin Pharmacol. 18:621-626

Martin W, Koselowske G, Töberich H, Kerkmann T, Mangold B and Augustin J （

1990) Pharmacokinetics and absolute bioavailability of ibuprofen after oral

administration of ibuprofen lysine in man. Biopharm Drug Dispos. 11:265-278.

McAllister RG Jr and Kirsten EB. (1982) The pharmacology of verapamil. IV.

Kinetic and dynamic effects after single intravenous and oral doses. Clin Pharmacol

Ther. 31:418-426


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

41

Moller A, Iwasaki K, Kawamura A, Teramura Y, Shiraga T, Hata T, Schafer A and

Undre NA. (1999) The disposition of 14C-labeled tacrolimus after intravenous and

oral administration in healthy human subjects. Drug Metabo. Dispos. 27:633-636

Naritomi Y, Terashita S, Kimura S, Suzuki A, Kagayama A and Sugiyama Y.

(2001) Prediction of human hepatic clearance from in vivo animal experiments and

in vitro metabolic studies with liver microsomes from animals and humans. Drug

Metab.Dispos. 29:1316-1324

Nishimura T, Amano N, Kubo Y, Ono M, Kato Y, Fujita H, Kimura Y and Tsuji A.

(2007) Asymmetric intestinal first-pass metabolism causes minimal oral

bioavailability of midazolam in cynomolgus monkey. Drug Metab Dispos.

35:1275-1284.

Patel RB, Patel UR, Rogge MC, Shah VP, Prasad VK, Selen A and Welling PG.

(1984) Bioavailability of hydrochlorothiazide from tablets and suspensions. J.

Pharm. Sci. 3:359-361

Pybus J and Bowers GN Jr. (1970) Measurement of serum lithium by atomic

absorption spectroscopy. Clin Chem. 16:139-143


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

42

Sakuda S, Akabane T and Teramura T. (2006) Marked species differences in the

bioavailability of midazolam in cynomolgus monkeys and humans. Xenobiotica.

36:331-40

Schulz P, Turner-Tamiyasu K, Smith G, Giacomini KM and Blaschke TF （1983)

Amitriptyline disposition in young and elderly normal men. Clin Pharmacol Ther.

33:360-366.

Shen DD, Kunze KL and Thummel KE. (1997) Enzyme-catalyzed processes of

first-pass hepatic and intestinal drug extraction. Adv. Drug Delivery Rev. 27:99-127

Shibata Y, Takahashi H, Chiba M and Ishii Y. (2002) Prediction of hepatic

clearance and availability by cryopreserved human hepatocytes: an application of

serum incubation method. Drug Metab Dispos. 30:892-896

Tabata K, Hamakawa N, Sanoh S, Terashita S and Teramura T. (2009) Exploratory

population pharmacokinetics (e-PPK) analysis for predicting human PK using

exploratory ADME data during early drug discovery research. Eur J Drug Metab

Pharmacokinet. 34:117-28

Takahashi M, Washio T, Suzuki N, Igeta K, Fujii Y, Hayashi M, Shirasaka Y and

Yamashita S. (2008) Characterization of gastrointestinal drug absorption in

cynomolgus monkeys. Mol. Pharm. 5:340-348


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

43

Tamura K, Kobayashi M, Hashimoto K, Kojima K, Nagase K, Iwasaki K, Kaizu T,

Tanaka H and Niwa M. (1987) A highly sensitive method to assay FK-506 levels in

plasma. Transplant Proc. 19(Suppl 6):23-29

Thummel KE, O'Shea D, Paine MF, Shen DD, Kunze KL, Perkins JD and

Wilkinson GR. (1996) Oral first-pass elimination of midazolam involves both

gastrointestinal and hepatic CYP3A-mediated metabolism. Clin Pharmacol Ther.

59:369-375

Uno Y, Hosaka S, Matsuno K, Nakamura C, Kito G, Kamataki T and Nagata R.

(2007) Characterization of cynomolgus monkey cytochrome P450 (CYP) cDNAs: is

CYP2C76 the only monkey-specific CYP gene responsible for species differences in

drug metabolism? Arch Biochem Biophys. 466:98-105

Walle T, Fagan TC, Conradi EC, Walle UK and Gaffney TE (1979) Presystemic and

systemic glucuronidation of propranolol. Clin Pharmacol Ther. 26:167-172

Wallemacq PE, Firdaous I, Hassoun A. (1993) Improvement and assessment of

enzyme-linked immunosorbent assay to detect low FK506 concentrations in plasma

or whole blood within 6 hours. Clin. Chem. 39:1045-1049


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

44

Wilson TW, Firor WB, Johnson GE, Holmes GI, Tsianco MC, Huber PB and Davies

RO. (1982) Timolol and propranolol: bioavailability, plasma concentrations, and

beta blockade. Clin.Pharmcol and Ther. 32:676-685

Wishart DS. (2007)Improving early drug discovery through ADME modeling: an

overview. Drug R D 8:349-362

Yang J, Jamei M, Yeo KR, Tucker GT and Rostami-Hodjegan A. (2007) Prediction

of intestinal first-pass drug metabolism. Curr Drug Metab. 8:676-684

Yang XX, Hu ZP, Duan W, Zhu YZ and Zhou SF. (2006) Drug-herb interactions:

eliminating toxicity with hard drug design. Curr Pharm Des. 12:4649-4664

Yu DK. (1999) The contribution of P-glycoprotein to pharmacokinetic drug-drug

interactions. J Clin. Pharmacol. 39:1203-1211


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

45

Legends for figures

Fig 1. Correlation of F, FaFg, and Fh in humans and cynomolgus monkeys. Open circle,

open triangle, open square, closed circle, and closed triangle represent category types A-E,

respectively.

Fig 2. Correlation of CLint intestine in humans and cynomolgus monkeys. NIF, MDZ, QID,

TAC, and VER represent nifedipine, midazolam, quinidine, tacrolimus, and verapamil

respectively.

Fig 3. Correlation of FaFg, and CLint liver in humans. Open triangle, open square, and

closed triangle represent category types B, C, and E respectively.


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


DMD 28829

46

TABLE 1.

Classification of each drug based on CYP isozyme selectivity and P-gp affinity

Type Compounds F in humans CYP isozyme P-gp affinity Reference

A Lithium 94.5% - - Arancibia et al., 1986

Hydrochlorothiazide 60.2% - - Patel et al., 1984

B

Dexamethasone 81.4% 3A4 ± Duggan et al., 1975

Nifedipine 41.2% 3A4 - Holtbecker et al., 1996

Midazolam 30.0% 3A4 ± Thummel et al., 1996

C

Quinidine 79.5% 3A4 + Greenblatt et al., 1977

Tacrolimus 23.3% 3A4 + Moller et al., 1999

Verapamil 18.0% 3A4 + McAllister and Kirsten, 1982

D Digoxin 65.3% - + Hinderling and Hartmann, 1991

E

Propranolol 29.0% 2D6,1A2 - Borgstrom et al., 1981

Amitriptyline 47.7% 2C19,2D6,3A4 ± Schulz et al., 1983

Timolol 61.0% 2D6 - Wilson et al., 1982

Ibuprofen 100% 2C9 - Martin et al., 1990

+: Drugs generally known to be P-gp substrates, ±: Drugs known to be P-gp substrates, -: Drugs generally not considered to be P-gp

substrates

This article has not been copyedited and form

atted. The final version m

ay differ from this version.

DM

D Fast Forw

ard. Published on Novem

ber 12, 2009 as DO

I: 10.1124/dmd.109.028829

at ASPET Journals on April 1, 2020 dmd.aspetjournals.org Downloaded from


DMD 28829

47

TABLE 2. Experimental conditions of cynomolgus monkey pharmacokinetic studies

Compounds Dosing route Dose Vehicle Volume Sample Points

mg/kg mL/kg h

Lithium Intravenous 5

Equivalent amount of hydrochloric acid 1 0.083, 0.25, 1, 3, 5, 8, 24

Oral 10 2 0.25, 0.5, 1, 2, 4, 8, 24

Hydrochlorothiazide Intravenous 1

50%PEG 1 0.25, 1, 1,5, 4, 6, 8, 24

Oral 1 2 0.5, 1.5, 2.5, 4, 6, 8, 24

Dexamethasone Intravenous 0.25

50%PEG 1 0.083, 0.25, 1, 2, 4, 6, 8

Oral 0.5 2 0.25, 0.5, 1, 2, 4, 6, 8

Nifedipine Intravenous 0.1

50%PEG 1 0.083, 0.25, 0.5, 1, 2, 4, 5

Oral 1 2 0.25, 0.5, 1, 2, 4, 5

Midazolam

Sakuda et. al., 2006

Intravenous 1 Distilled water a

1 0.1, 0.25, 0.5, 1, 2, 4, 8, 12, 24

Oral 3 2 0.25, 0.5, 1, 2, 4, 8, 12, 24

Quinidine Intravenous 1 Saline 1 0.1, 0.25, 0.5, 1, 2, 4, 8, 24

Oral 3 Distilled water 2 0.25, 0.5, 1, 2, 4, 8, 24

Tacrolimus Intravenous 0.004

Saline b 0.5 0.083, 0.25, 0.5, 1, 4, 8, 24

Oral 0.02 2 0.25, 0.5, 1, 2, 4, 8, 24

Verapamil Intravenous 1

Saline 1 0.1, 0.25, 0.5, 1, 2, 4, 8, 12, 24

Oral 3 2 0.25, 0.5, 1, 2, 4, 8, 12, 24

Digoxin Intravenous 0.1

50%PEG 1 0.1, 0.25, 0.5, 1, 2, 4, 8, 24

Oral 0.1 2 0.25, 0.5, 1, 2, 4, 8, 24

Propranolol Intravenous 0.3 Saline 2 0.1, 0.25, 0.5, 1, 2, 4, 8, 24


Amitriptyline Intravenous 0.3 Saline 2 0.1, 0.25, 0.5, 1, 2, 4, 8, 24


Timolol Intravenous 0.3 Saline 2 0.1, 0.25, 0.5, 1, 2, 4, 8, 24





DM

D Fast Forw


ber 12, 2009 as DO

I: 10.1124/dmd.109.028829



DMD 28829

48

Ibuprofen Intravenous 1 Saline 2 0.1, 0.25, 0.5, 1, 2, 4, 8, 24

Oral 3 50%PEG 2 0.25, 0.5, 1, 2, 4, 8, 24

The intravenous and oral administrations conditions for MDZ were taken from Sakuda et. al., 2005. a Dormicam was diluted with distilled water b Prograf was diluted with saline




DM

D Fast Forw


ber 12, 2009 as DO

I: 10.1124/dmd.109.028829



DMD 28829

49

TABLE 3. Apparatus and LC/MS/MS analytical conditions for each drug during the determination of concentration in cynomolgus

monkey plasma

LC

system

MS/MS

system Column

Column

temperature

Injection

volume

Flow

rate Mobile Phase

°C uL mL/min

Hydrochlorothiazide HP1100 API-2000 Inertsil ODS 3.3 uM (2.1 x 50 mm) 40 40 0.2 0.1% Formic acid :

Acetonitrile = 1:1

Nifedipine HP1100 API-2000 Inertsil ODS 3.5 uM (3.0 x 150 mm) RT 30 0.2 Water : Acetonitrile = 4:6

Quinidine LC-VP/LC-10A API-3000 Xterra MS C18 (4.6 x 50 mm) 40 10 0.3 20 mM Ammonium acetate

(pH4.8) : Acetonitrile = 4:6

Verapamil LC-VP/LC-10A API-3000 Xterra MS C18 (4.6 x 50 mm) 40 10 0.3 20 mM Ammonium acetate

(pH4.8) : Acetonitrile = 4:6

Digoxin LC-VP/LC-10A API-3000 Xterra MS C18 (4.6 x 30 mm) 40 20 0.3 2 mM Ammonium acetate :

Acetonitrile =65:35

Propranolol LC-VP/LC-10A API-3000 Xterra MS C18 (4.6 x 50 mm) 40 10 0.3 20 mM Ammonium acetate

(pH4.8): Acetonitrile = 4:6

Amitriptyline LC-VP/LC-10A API-3000 Xterra MS C18 (4.6 x 50 mm) 40 10 0.3 20 mM Ammonium acetate


Timolol LC-VP/LC-10A API-3000 Xterra MS C18 (4.6 x 50 mm) 40 10 0.3 20 mM Ammonium acetate


Ibuprofen LC-VP/LC-10A API-3000 Xterra MS C18 (4.6 x 50 mm) 40 10 0.3 20 mM Ammonium acetate





DM

D Fast Forw


ber 12, 2009 as DO

I: 10.1124/dmd.109.028829



DMD 28829

50

TABLE.4. Summary of in vivo pharmacokinetic parameter values in humans and cynomolgus monkeys

Drugs Species Dose (i.v./p.o.) CLt fe CLh F FaFg Fh

mg/kg mL/min/kg % mL/min/kg %

Lithium human (-/0.25) 0.4±0.2 No Data 0a 94.5±15.8 0.95 1a)

monkey 0.14/0.27 0.7±0.1 No Data 0a 97.9±6.8 0.98 1a)

Hydrochlorothiazide human (-/0.32) 3.0±1.0 60.2 0a 60.2 0.60 1a)

monkey 1/1 5.9±2.0 No data 0a 30.7±9.4 0.31 1a)

Dexamethasone human 0.17/0.17 2.7±0.8 10.8±4.3 2.4 81.4±15.8 0.93 0.88

monkey 0.25/0.5 4.5±0.8 No data 4.5 78.9±9.8 0.85 0.93

Nifedipine human 0.02/0.27 8.2±0.6 No data 8.2 41.2±5.4 0.89 0.46

monkey 0.1/1 15.6±3.5 0.057 15.6 9.3±4.0 0.19 0.48

Midazolam human 0.013/0.026 4.7±1.5 0.27±0.07 4.7 30.0±10.0 0.45 0.67

monkey 1/3 12.9± 1.8 <1% 12.9 2.0±0.4 0.03 0.62

Quinidine human 4.3/5.0 3.8±0.3 35.1±1.8 2.5 79.5±15.0 0.96 0.83

monkey 1/3 12.8±0.7 0.6±0.2 12.7 4.5±1.7 0.07 0.62

Tacrolimus human 0.02/0.05 0.5±0.1 0.04±0.02 0.5 23.3±16.7 0.24 0.98

monkey 0.004/0.02 2.6±0.3 No data 2.6 0.5±0.5 0.005 0.94

Verapamil human 0.14/1.14 11.8±0.5 No data 11.8 18.0±10.1 0.47 0.38

monkey 1/3 44.9±10.5 1.5±0.7 44.2 0 -b) 0.00

Digoxin human 0.01/0.01 2.9±0.6 80.5±3.2 0.6 65.3±22.5 0.67 0.97

monkey 0.1/0.1 2.9±0.03 17.1±9.3 2.4 45.0±14.0 0.48 0.94

Propranolol human 0.13/0.5 11.6c) No data 11.6 29 0.78 0.37

monkey 0.3/1 24.3±2.4 No data 24.3 3.3±1.5 0.10 0.34

Amitriptyline human 0.6/1.2 12.5±2.3 No data 12.5 47.7±11.0 1d) 0.30

monkey 0.3/1 35.8±8.8 0.2±0.2 35.7 1.3±1.0 0.03 0.41

Timolol human 0.025/0.4 7.7±3.7 No data 7.7 61.0±19.2 1d) 0.56

monkey 0.3/1 13.6±0.4 4.8±2.6 13 10.8±4.3 0.15 0.71




DM

D Fast Forw


ber 12, 2009 as DO

I: 10.1124/dmd.109.028829



DMD 28829

51

Ibuprofen human 2.9/4.2 0.8±0.2 No data 0.8 102.8±12.0 1d) 0.93

monkey 1/3 7.9±0.7 18.5±1.1 6.4 103.4±14.2 1d) 0.76

Mean values were used for CLh, F, and FaFg.

Li and HT were assumed that fe after oral administration was equal to F.

MDZ in cynomolgus monkey were taken from Sakuda et. al., 2005. a) Assuming the Fh values were 1. b) Not calculated. c) CLt was calculated by dividing dose by AUC after intravenous administration.. d) The calculated values were greater than 1.




DM

D Fast Forw


ber 12, 2009 as DO

I: 10.1124/dmd.109.028829



DMD 28829

52

TABLE 5. Summary of in vitro pharmacokinetic parameters of all tested drugs in humans and cynomolgus monkeys

Drugs Species Rb Papp Protein binding fp fb CLint liver CLint intestine fb*CLint liver/Qh ATPase

×10-6cm/sec % mL/min/kg uL/min/mg protein ratio vs VER

Hydrochlorothiazide human 2.70

0.1 40.0 0.600 0.222 - a - a - a 0

monkey 1.84 39.0 0.610 0.331 - a - a - a NT

Dexamethasone human 0.95

16.4 52.0 0.480 0.507 66 - a 1.62 0.01

monkey 1.34 77.5 0.225 0.167 24 - a 0.09 NT

Nifedipine human 0.74

13.4 93.4 0.066 0.089 438 138 1.89 0.03

monkey 0.65 94.3 0.057 0.088 2597 612 4.96 NT

Midazolam human 0.69

30.8 97.0 0.030 0.044 877 385 1.85 0.28

monkey 0.77 95.7 0.043 0.056 1422 1635 1.72 NT

Quinidine human 0.72

17.9 91.4 0.086 0.119 52 - a 0.30 4.57

monkey 0.78 92.1 0.079 0.102 872 212 1.91 NT

Tacrolimus human 20.00

34.2 98.9 0.011 0.001 1538 625 0.04 7.89

monkey 20.00 99.0 0.010 0.001 5793 4663 0.06 NT

Verapamil human 0.92

35.8 95.2 0.048 0.052 656 69 1.65 1.00

monkey 0.93 88.2 0.118 0.127 2491 696 6.83 NT

Digoxin human 1.00

0.1 60.3 0.397 0.398 - a - a - a 56.1

monkey 0.82 52.9 0.471 0.574 - a - a - a NT

Propranolol human 0.89 b

37.4 86 0.140 0.157 165 - a 1.25 1.70

monkey 0.85 c 78.8 0.212 0.249 974 - a 5.25 NT

Amitriptyline human 0.86 d

53.3 85.4 0.146 0.170 80 - a 0.66 1.20

monkey 1.40 87.2 0.128 0.091 2559 - a 5.06 NT

Timolol human 0.84 b

27.3 50.9 0.491 0.585 32 - a 0.89 1.66

monkey 1.02 95.5 0.045 0.044 391 - a 0.37 NT

Ibuprofen human 0.55 d

29.4 98.8 0.012 0.022 38 - a 0.04 1.14

monkey 0.61 98.5 0.015 0.024 25 - a 0.01 NT




DM

D Fast Forw


ber 12, 2009 as DO

I: 10.1124/dmd.109.028829



DMD 28829

53

Lithium was excluded from all in vitro studies.

NT: Not tested a The CLint couldn’t be calculated because the tested drug was not depleted.

b Data were taken from Shibata et. al., 2002. c Data were taken from Evans et. al., 1973. d Data were taken from Obach et. al., 2005.




DM

D Fast Forw


ber 12, 2009 as DO

I: 10.1124/dmd.109.028829



Fig. 1F

0%

20%

40%

60%

80%

100%

0% 20% 40% 60% 80% 100%

Human

Monkey

Fh

0%

20%

40%

60%

80%

100%

0% 20% 40% 60% 80% 100%

Human

Monkey

FaFg

0%

20%

40%

60%

80%

100%

0% 20% 40% 60% 80% 100%

Human

Monkey


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


Fig. 2

TAC

MDZ

NIF QID

VER

CLint.intestine

0

1000

2000

3000

4000

5000

0 1000 2000 3000 4000 5000

Human (uL/min/mg protein)

Monkey

(uL/min/mg protein)


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


Fig. 3

Human

0%

20%

40%

60%

80%

100%

0 400 800 1200 1600

CLintliver

(mL/min/kg)

FaFg


at ASPE

T Journals on A

pril 1, 2020dm


ownloaded from


Documents

DMD 28829 A Comparison of Pharmacokinetics Between …dmd.aspetjournals.org/content/dmd/early/2009/11/12/dmd.109.028829.full.pdfpredict human pharmacokinetics. Chou et al., reported